Download Polycystin-2 functions as an intracellular calcium release channel.

articles
Polycystin-2 is an intracellular
calcium release channel
Peter Koulen*†#, Yiqiang Cai‡**, Lin Geng‡**, Yoshiko Maeda‡, Sayoko Nishimura‡, Ralph Witzgall§,
Barbara E. Ehrlich*† and Stefan Somlo‡||¶
Departments of *Pharmacology, †Cellular & Molecular Physiology, ‡Internal Medicine and ¶Genetics, Yale University School of Medicine, 333 Cedar Street, New
Haven, Connecticut 06520, USA
§Institute of Anatomy and Cell Biology, University of Heidelberg, 69120, Heidelberg, Germany.
#Present address: Department of Pharmacology and Neuroscience, University of North Texas Health Science, 3500 Camp Bowie Boulevard, Fort Worth, Texas,
76107, USA
||e-mail: [email protected]
**These authors contributed equally to this work and are listed alphabetically..
Published online: 18 February 2002, DOI: 10.1038/ncb754
Polycystin-2, the product of the gene mutated in type 2 autosomal dominant polycystic kidney disease (ADPKD), is
the prototypical member of a subfamily of the transient receptor potential (TRP) channel superfamily, which is
expressed abundantly in the endoplasmic reticulum (ER) membrane. Here, we show by single channel studies that
polycystin-2 behaves as a calcium-activated, high conductance ER channel that is permeable to divalent cations.
Epithelial cells overexpressing polycystin-2 show markedly augmented intracellular calcium release signals that are
lost after carboxy-terminal truncation or by the introduction of a disease-causing missense mutation. These data suggest that polycystin-2 functions as a calcium-activated intracellular calcium release channel in vivo and that polycystic kidney disease results from the loss of a regulated intracellular calcium release signalling mechanism.
DPKD affects more than 1 in 1000 live births and is the most
common monogenic cause of kidney failure in man1.
Mutations in either of two genes, PKD1 (ref. 2) or PKD2 (ref.
3), cause virtually indistinguishable clinical presentations; most
notably, growth of fluid-filled cysts in the kidney, liver and pancreas. The primary molecular genetic basis for cyst formation is
homozygous loss-of-function because of somatic second hits in the
setting of inherited inactivating mutations4,5. The near identity of
polycystic disease symptoms, irrespective of the causative gene,
suggests that polycystin-1 and -2 function along a common pathway. This hypothesis is supported by a variety of observations,
including the similarity of phenotypes between mouse knockouts
of either gene6,7, the discovery that the two proteins interact8,9 and
the identity of the phenotypes caused by mutations in homologous
genes in Caenorhabditis elegans10.
Polycystin-2, the product of the gene mutated in type 2 ADPKD
(online MIM number, 173910)3 is the prototypical member of a
subfamily of the TRP channel superfamily11. Immunocytochemical
colocalization, subcellular fractionation and endoglycosidase H
(Endo H) sensitivity analyses have established that epithelial cells
express polycystin-2 exclusively in pre-medial Golgi membranes,
most notably the ER9,12. In addition, the cytoplasmic tail of polycystin-2 contains signals that are necessary for ER retention9,12.
Cell-surface biotinylation assays also suggest that native polycystin2 is not transported to the cell surface in epithelial cells12.
Immunohistochemical studies have defined the pattern of polycystin-2 tissue-specific expression13–15, but ultrastructural studies to
define the subcellular location of the protein in tissues have not, as
yet, been reported. Furthermore, recent studies of cation channel
activities attributed to polycystin-2 in heterologous systems have
been unable to determine the site or mechanism of action of the
protein in epithelial cells9,16,17.
In the present study, we examined both the mechanism and site
of action of polycystin-2. Single channel recordings from ER
microsomes fused to lipid bilayers show that polycystin-2 is a high
A
conductance channel that is permeable to divalent cations. Calcium
imaging studies in epithelial cells over-expressing polycystin-2
demonstrate enhanced inositol 1,4,5-trisphosphate (InsP3)-mediated calcium release from intracellular pools. Mutation of polycystin-2
abrogates this calcium release response. Gradient fractionation and
Endo H sensitivity analyses of native kidney tissue show that the distribution of polycystin-2 is indistinguishable from that of ER membrane proteins. The data suggest that polycystin-2 is highly expressed
in the ER of kidney epithelial cells and functions as a calcium release
channel.
Results
Single channel properties of polycystin-2. LLC-PK1 porcine kidney
cell lines stably expressing either full-length PKD2 (ref. 12), a Cterminal truncated polycystin-2 mutant (PKD2-L703X) (ref. 12),
or a naturally occurring missense variant of polycystin-2 (PKD2D511V)18 were used to study the channel properties of the protein
(Fig. 1a; see Methods). PKD2-L703X, which lacks several putative
protein interaction domains3,8,12,19, is representative of clinically relevant premature termination variants (Fig. 1a). Full-length polycystin-2 is confined to pre-medial Golgi membranes, whereas the
PKD2-L703X mutant partially translocates to the plasma membrane because of the loss of essential ER retention signals12. The
D511V mutant is a pathogenic missense variant that differs from
the wild-type sequence by a single amino acid in the third membrane-spanning domain18 (Fig. 1a). As the D511V mutant has an
intact C terminus, we would predict that it will retain the cellular
distribution of the wild-type protein. An immunocytochemical
examination of LLC-PK1 cells stably overexpressing PKD2-D511V
with YCC2 antisera5,12,13,15,18–20, revealed a pattern of expression that
was indistinguishable from the wild-type polycystin-2 pattern
(Fig. 1b), and consistent with abundant expression in ER membranes.
A semiquantitative assessment of PKD2 and PKD2-D511V overexpression by serial dilution and immunoblotting demonstrated that
NATURE CELL BIOLOGY ADVANCE ONLINE PUBLICATION http://cellbio.nature.com
© 2002 Macmillan Magazines Ltd
1
articles
a
1
Wt
23 4 5 6
c
PKD2
1
EF
2 3 4 5 6 hand
PKD2-L703X
PKD2-D511V
1x
PKD2
1x
10x 20x 30x 40x 50x 100x Serial dilution (fold)
5x
Cell
lysate
PKD2
1
23 4 5 6
D511V
PKD2
EF
hand
1x
2x
Wt Vector
1x Serial dilution (fold)
5x 10x 20x 30x 50x 1x
Microsomes
PKD2
b
PKD2
PKD2-D511V
d
D511V
1x
10x
Wt
D511V
Wt
1x
1x
1x
10x
Serial dilution (fold)
PKD2-D511V
Calnexin
PKD2-D511V
Microsomes
Calnexin
Cell lysate
e
PKD2
Untransfected
–
1x
– +
PKD2
10x
1x
Doxycycline
Serial dilution (fold)
LtTA-2.22
lysate
Calnexin
Pre-immune
Figure 1 Expression of polycystin-2 in cell lines. a, Schematic representation of
full length PKD2 and the PKD2-L703X and PKD2-D511V mutants. PKD2-L703X
lacks the last C-terminal 265 amino acids, PKD2-D511V has a missense change in
the third transmembrane-spanning domain. Dark boxes (1–6) represent membranespanning domains; shaded box represents the EF-hand. b, Confocal immunofluorescence images of LLC-PK1 cells stained with YCC2 antiserum, showing that overexpressed PKD2 (top left) and PKD2-D511V (top right) have identical cytoplasmic
reticular staining patterns. Untransfected parental cells (lower left) and PKD2-D511V
over-expressing cells (lower right) were stained with YCC2 and pre-immune serum,
respectively. c, Semiquantitative analysis of protein expression by serial dilution
shows a 5–10-fold enrichment of PKD2 expression in total lysate and microsomes
from overexpressing cells. 1x represents 15 µg total protein loaded in this experiment, and also in d and e. Serial dilutions of the 15-µg samples (2–100x) were
made thereof, and also apply to d and e. Wt refers to non-overexpressing cells
throughout. d, Approximately 10-fold overexpression of PKD2-D511V in cell lysates
and microsomes from the PKD2-D511V-expressing cell line. Calnexin is used as
internal loading control. e, The inducible LtTA-2.22 cell line achieves ~10-fold overexpression of polycystin-2. Cells were incubated in the absence (−) or presence (+)
of doxycycline. The same blot was also reprobed with an anti-calnexin antibody (bottom panel).
the proteins are enriched by between 5- and 10-fold, respectively, in
the stable cell lines (Fig. 1c,d). An LLC-PK1 cell line (LtTA-2.22)
that inducibly over-expresses wild-type polycystin-2 also achieved
~10-fold overexpression in the induced state (Fig. 1e). The
immunostaining pattern of polycystin-2 in LtTA-2.22 cells was
indistinguishable from that in stable, non-inducible, overexpressing cells (data not shown).
To investigate the function of polycystin-2 in vitro, we began by
defining its channel properties and structure-function relationships in lipid bilayers. The presence of ER membranes in microsome fractions was examined by immunoblotting with antisera to
the ER-resident protein calnexin (Fig. 1d,e). The successful fusion
of microsomes to lipid bilayers was documented by recording large
potassium and chloride ion channel currents (Fig. 2a, traces A,C),
which disappeared after removal of potassium chloride from the cis
side (the cytoplasmic side of the channel) of the bilayer (Fig. 2a,
traces B, D). To monitor currents in the absence of potassium chloride, barium ions on the trans side of the bilayer (ER luminal side of
the channel) were used as a current carrier. Inward barium currents
were detected in ER vesicle preparations that overexpressed polycystin-2 (n = 24/24 channels from 8 independent ER microsome
preparations; Fig. 2a, trace B). By contrast, no such currents were
observed when ER vesicle preparations from wild-type LLC-PK1
cells (n = 0/47 channels from 3 ER preparations; Fig. 2a, trace D) or
LLC-PK1 cells expressing the empty vector alone (n = 0/7 channels
from 2 ER preparations, data not shown) were fused to bilayers. The
possibility that the currents detected here were mediated by InsP3
receptor (InsP3R) channels was excluded by the absence of InsP3 in
these experiments. Any potential contribution of ryanodine receptor (RyR) channels to the current was also ruled out by the addition
of the RyR inhibitor ruthenium red (2 µM), to the cytoplasmic side
of the membrane during the entire course of the experiment.
Channel activity under these conditions indicates the presence of a
novel intracellular calcium release channel—namely, polycystin-2.
Larger negative holding potentials increased the current amplitudes for wild-type polycystin-2 (Fig. 2b). The polycystin-2 channel had slope conductances of 114, 95 and 90 pS when measured
between 0 and −10 mV with barium, calcium and magnesium as
the respective current carriers (Fig. 2b). Barium was used as the
current carrier for the majority of subsequent bilayer experiments
because it does not influence the activity of known intracellular calcium channels and generates larger current amplitudes for polycystin-2 (barium, 2 pA; calcium, 1.7 pA; magnesium, 1.5 pA; all
measured at −10 mV). These electrophysiological properties of
polycystin-2 parallel those described previously for InsP3R and
RyR21–23, as well as other TRP-related channels24, but differ from
voltage-operated channels at the cell surface25. The PKD2-L703X
truncation mutant had reduced current amplitudes and required
larger negative membrane potentials for channel activation than
the wild-type protein (n = 18/18 channels from 4 ER microsome
preparations; Fig. 2c). The slope conductance was reduced to 28 pS
between −20 and −60 mV with barium as the current carrier
(Fig. 2c). When ER vesicles from cells expressing the PKD2-D511V
mutant were fused to bilayers, no currents were observed (n = 0/12
2
NATURE CELL BIOLOGY ADVANCE ONLINE PUBLICATION http://cellbio.nature.com
© 2002 Macmillan Magazines Ltd
articles
a
A
B
C
D
5 pA
Holding potential (mV)
–40 –30 –20 –10
–0
10
20
0
–0.5
–1
–1.5
–2
–2.5
Ba2+
0.6
0.4
0.2
0
0.001
0.1
10
1000
Ca2+ concentration (µM)
b
Amplitude (pA)
b
0.8
1
0 mV
Ca2+
0.4
–20 mV
250 ms
0.2
100 ms 1 pA
1 pA
0
–30 –20 –10 0 10 20
Holding potential (mV)
c
0
–0.5
–1
–1.5
–2
–2.5
–3
c
Amplitude (pA)
Holding potential (mV)
–50 –25
0
25 50
0.6
Mg2+
–3
–60
0.8
–5 mV
1
0 mV
0.8
–20 mV
0.6
0.4
–60 mV
0.2
Open probability
(normalized)
250 ms
1
Open probability
(normalized)
Open probability (normalized)
a
100 ms 1 pA
0
Figure 2 Polycystin-2 is a voltage-dependent ion channel that conducts
divalent cations. a, ER-derived vesicles from LLC-PK1 cells over-expressing polycystin-2 (A, B) or wild type LLC-PK1 cells (C, D) were fused to lipid bilayers. Fusion
was documented by the presence of potassium and chloride channels (A, C). After
removal of potassium chloride from the medium, single channel currents generated
by polycystin-2 were recorded with 55 mM barium as the sole current carrier, at a
holding potential of −15 mV (B). Channel activity was not observed with ER vesicles
from nontransfected cells (D). Channel openings are shown as downward deflections and bars to the right indicate zero current. b, Channel current amplitude-voltage relationships of polycystin-2, recorded with 55 mM barium (open square;
n = 24), calcium (filled circle; n = 7) or magnesium (filled square; n = 5) as current
carriers (left panel). Slope conductances, determined by linear regression between
0 and −10 mV, were 114 pS, 95 pS and 90 pS, respectively. Single channel currents recorded at −10 mV show that polycystin-2 conducts barium (average amplitude, 2 pA), calcium (average amplitude, 1.7 pA), and magnesium (average amplitude, 1.5 pA), (right panel). c, The channel current amplitude-voltage relationship of
PKD2-L703X generated a slope conductance of 28 pS, determined between −20
and −60 mV.
channels from 3 ER preparations; data not shown). The latter finding provides evidence that this missense mutation in polycystin-2
results in loss of function. Coupled with the pathogenicity of this
variant in patients18, the data support the hypothesis that the channel activity of polycystin-2, independent of protein interactions in
the C terminus, is required for the normal functioning of the polycystin signalling pathway in the kidney.
Given the effects of intracellular calcium on the activity of intracellular calcium channels26, we examined polycystin-2 channel
activity in the presence of varying cytosolic calcium concentrations. The normalized open probability of polycystin-2 increased
from 0.19 to 1.0 over a range of free calcium concentrations from
0.01 to 1260 µM (Fig. 3a). These concentrations include the physiological range (0.1–10 µM) of free cytosolic calcium. Higher free
cytosolic calcium concentrations lowered the open probability
(Fig. 3a). The PKD2-L703X mutant was not activated over the range
of calcium concentrations tested (Fig. 3a). Whereas physiological
–60
–40 –20
0
20
Holding potential (mV)
Figure 3 Dependence of polycystin-2 channel open probability on intracellular free calcium concentration and membrane holding potential. a, The
open probabilities of PKD2 channels (open square; 0 mV; n = 5) and PKD2-L703X
channels (filled circle; −50 mV; n = 7) were monitored at different free calcium concentrations on the cytosolic side of the channel. The PKD2 channel was activated
by submillimolar and millimolar calcium concentrations on its cytosolic side (open
probability at 100 µM, 3.3%±1.3% (S.E.M; n = 5). Higher cytosolic free calcium
inhibited channel activity. Varying cytosolic free calcium concentrations did not
effect the PKD2-L703X channel. Data were normalized to the maximum open probability. b, Single channel recordings of PKD2 activity (left panel) at 0 mV (Popen,
0.2%), −5 mV (Popen, 1.1%) and –20 mV (Popen, 2.5%). These open probabilities are
for the experiment shown in the panel; the current carrier is barium. The open
probability of the PKD2 channel was monitored at different membrane holding
potentials (n = 24; right panel). Data were normalized to the open probability at −
30 mV. The mean maximal open probability at −30 mV was 3.0%±0.2% (S.E.M);
n = 24). c, Typical recordings, as in b, for the PKD2-L703X channel activity (left
panel) at holding potentials of 0 mV (Popen, 0%), −20 mV (Popen, 0.1%) and −60 mV
(Popen, 2.4%) and as a function of the holding potentials (n = 18; right panel). The
open probabilities of PKD2-L703X channel were similar to the wild-type channel, but
were only achieved when the membrane potential was significantly larger. Data
were normalized to the open probability at −60 mV.
concentrations of free calcium enhance the activity of polycystin-2
channels, concentrations outside this range induce channel inhibition. The absence of such modulation in the PKD2-L703X mutant
suggests that this functional regulation may be mediated by sites
located in the cytoplasmic tail of polycystin-2 (ref. 27). Increasingly
negative holding potentials also increased the activity of polycystin2 channels, demonstrating that the channel is voltage-dependent
(Fig. 3b). The maximum open probability was achieved at holding
potentials of −20 mV or greater. The mean maximal open probability at −30 mV was 3.0% (n = 24, ±0.2% S.E.M). This is comparable
NATURE CELL BIOLOGY ADVANCE ONLINE PUBLICATION http://cellbio.nature.com
© 2002 Macmillan Magazines Ltd
3
articles
b
c
1.5
Vector
PKD2-L703X
1
21
PKD2-D511V
0.5
0
60
120
300
250
200
150
100
50
0
CP
Ve K
cto
D2 r
D2
-W
-W
t-[C t
2
LtT LtT a +]
A
A2.2 -2.2
2
+ 2
PK
D2 doxy
PK -L70
D2
3
-D X
51
1V
PKD2-Wt
2
180
PK
Duration of Ca2+ transient (s)
CP
Ve K
cto
r
D
D2
-W 2-W
t-[C t
2+
a
LtT LtT
]
A
A2.2 -2.2
2
PK 2 +do
D2
xy
PK -L70
3
D2
-D X
51
1V
PK
3.5
3
2.5
2
1.5
1
0.5
0
e
300
250
200
150
100
50
0
0 10 50 100 200
2-APB concentration (µM)
PK
63
d
LL
42
Amplitude of Ca2+ transient (ratio F/F0)
Time (s)
PK
0
Vector
LL
Time (s)
PKD2-Wt
Ratio F/F0
a
Duration of Ca2+ transient (s)
2.5
Figure 4 Overexpression of polycystin-2 enhances amplitude and duration of
vasopressin-induced calcium transients. a, LLC-PK1 cells stably transfected
with polycystin-2 (PKD2-Wt) or the empty expression plasmid (Vector) were loaded
with the fluo-3 dye. All cells respond to a vasopressin-stimulus (70 nM, applied at 0
s) with an increase in intracellular calcium, visualized by increasingly bright pixels
from the fluorescence of the indicator dye with increasing time. Overexpression of
PKD2 results in significantly larger calcium transients, compared to control cells.
Scale bar, 25 µm. b, Vasopressin treatment (70 nM) -induces single-peak increases
in cytosolic calcium. Changes in the ratio of calcium-dependent fluorescence over
prestimulus background fluorescence (F/F0) are plotted over time in single representative LLC-PK1 cells stably expressing full-length polycystin-2 (PKD2-Wt), the
empty expression plasmid (Vector) and the mutants PKD2-L703X and PKD2-D511V.
c, The mean duration (± S.E.M.) of vasopressin-induced calcium transients was
measured in untransfected LLC-PK1 cells (LLC-PK1), and in LLC-PK1 cells stably
expressing the empty expression plasmid (Vector), polycystin-2 (PKD2-Wt), polycystin-2 in calcium-free medium (PKD2-Wt-(Ca2+)o), a ‘tet-off’ inducible polycystin-2
incubated without (LtTA-2.22) and with doxycyclin (LtTA-2.22 +doxy), PKD2-L703X,
and PKD2-D511V, respectively. n = 50 for each condition. d, The mean amplitude
(± S.E.M.) of vasopressin-induced calcium transients measured in the same group
of LLC-PK1 cell lines. e, Vasopressin-induced calcium transients in cells overexpressing polycystin-2 can be blocked by 2-APB, in a dose-dependent manner. Mean
duration (± S.E.M.) of calcium transients in untransfected LLC-PK1 cells (grey bars;
n = 50) and in LLC-PK1 cells overexpressing polycystin-2 (white bars; n = 50).
to the open probability of ~4% detected for InsP3R in bilayers22, but
much lower than that measured for RyR21. Furthermore, variations
in membrane potential do not significantly alter the channel open
probability of either InsP3R or RyR23,28,29. No significant channel
activity was detected for the PKD2-L703X mutant at holding potentials up to −20 mV, which are sufficient to fully activate wild-type
polycystin-2 channels (Fig. 3c). This change in the channel properties of the PKD2-L703X mutant indicates that the C-terminal
domain is important in the normal channel function of polycystin2. Pathogenic premature termination mutations in polycystin-2, if
they produce stable proteins, result in channels with an altered single channel conductance and open probability, in addition to the
previously described effects on protein interactions8,9 and alterations in subcellular localization12.
Release of intracellular calcium by polycystin-2. The discovery that
polycystin-2 is expressed in the ER of epithelial cells and that it
functions as a calcium-activated calcium channel in lipid bilayers
led us to hypothesize that small changes in local intracellular calcium concentration may be sufficient to activate nearby polycystin-2
channels. To test this hypothesis, LLC-PK1 cells stably co-expressing
both wild-type and mutant polycystin-2 were stimulated with
vasopressin, to induce intracellular calcium release through
InsP3R30. LLC-PK1 cells express both vasopressin V1 and V2 receptors, but only V1 receptor stimulation, through activation of
InsP3R, results in increased levels of cytosolic calcium29,30.
Activation of V2 receptors stimulates the production of cAMP,
without inducing changes in the levels of intracellular calcium29,30.
LLC-PK1 cells, loaded with the calcium-sensitive fluorescent dye
fluo-3 acetoxymethyl ester (fluo-3) and expressing only endogenous polycystin-2, responded to vasopressin stimulation with a single transient increase in cytosolic calcium that declined to baseline
levels within 60 s (Fig. 4a,b)30. Similar results were also obtained
with LLC-PK1 cells that had been stably transfected with the empty
expression vector (Fig. 4a,b). By contrast, LLC-PK1 cells overexpressing full-length polycystin-2 generated an approximately two-fold
larger increase in cytosolic calcium and an approximately ten-fold
longer duration of the calcium transients (Fig. 4a–d). The increased
duration and amplitude of transients persisted in the absence of
extracellular calcium, suggesting that the increase in cytosolic calcium results from release of intracellular stores (Fig. 4c,d). We confirmed that InsP3R activation is required for polycystin-2-mediated
calcium release in our system by treating cells with the InsP3R
inhibitor 2-aminoethyl diphenyl borate (2-APB) (Fig. 4e).
Furthermore, we confirmed that 2-APB does not directly inhibit
polycystin-2 channel activity in bilayers (data not shown). We also
confirmed the integrity of intracellular calcium stores, and excluded the possibility that variations in the calcium content of the ER in
different cell lines accounted for the enhanced calcium transients,
4
NATURE CELL BIOLOGY ADVANCE ONLINE PUBLICATION http://cellbio.nature.com
© 2002 Macmillan Magazines Ltd
articles
Gradient fractions
a
1 2
3 4 5 6 7 8
9 10 11 12 13 14 15 16 17 18
Polycystin-2
Calnexin
NHE3
EGFR
Golgi 58K
Gradient fractions
b
1
2
3
4
5
6
7
8
9
– + – + – + – + – + – + – + – + – + Endo H
Gradient fractions
10
11
12
13
14
15
16
17
18
– + – + – + – + – + – + – + – + – + Endo H
5
c
–
+ Endo H
Figure 5 Polycystin-2 is expressed in the ER of native kidney. a, Kidney tissue
lysates were fractionated by density gradient centrifugation on linear 8–34%
Optiprep gradients. Equal volumes of the 18 fractions were loaded in each lane;
fraction 1 is the top of the gradient, fraction 18 is the bottom. Samples were
analysed by immunoblotting, with antibodies to polycystin-2, calnexin, NHE3, EGFR
and the Golgi 58K protein. b, Equal aliquots of each fraction was digested with
Endo H (+) or incubated with digestion buffer but no enzyme (−). All fractions are
sensitive to Endo H; fractions 4–6 show partial Endo H resistance (see text for
details). c, Polycystin-2 in fraction 5 is completely Endo H sensitive (+) after
removal of the lower molecular weight proteins.
by showing that the addition of thapsigargin releases similar
amounts of calcium in all of the cell lines tested (data not shown).
We next used the ‘tet-off ’ inducible polycystin-2-overexpressing
LLC-PK1 cell line (LtTA-2.22) to establish that the enhanced calcium transients are directly attributable to overexpression of polycystin-2. Inducible overexpression of polycystin-2 resulted in calcium transient amplitudes and durations that were indistinguishable
from polycystin-2 that had been overexpressed constitutively
(Fig. 4c,d). Suppression of polycystin-2 overexpression, in turn,
resulted in calcium transients of comparable amplitude and duration to transients detected in cells that do not overexpress functional polycystin-2 (Fig. 4c,d). LLC-PK1 cells that expressed either
the PKD2-L703X or PKD2-D511V mutants generated calcium
transients with amplitudes and durations that were similar to wild
type cells (Fig. 4c,d). The absence of enhanced calcium transients
with the truncated PKD2-L703X suggests that preservation
domains in the C terminus are required for the release of calcium
from intracellular stores. On the other hand, the absence of
enhanced calcium transients in the PKD2-D511V mutant provides
evidence that polycystin-2 mediates intracellular calcium release
through its own channel activity, rather than through an interaction with other channel proteins, such as InsP3R. These findings
support the hypotheses that polycystin-2 is an intracellular calcium
release channel that can be activated in epithelial cells by local
increases in calcium concentration.
Polycystin-2 is an ER protein in native kidney. The final arbiter of
the primary subcellular site of polycystin-2 action is the location of
the protein in the tubular epithelial cells of native kidney tissue. As
polycystin-2 and critical binding partners such as polycystin-1 are
expressed and functional at physiological levels in native kidney, we
reasoned that determination of the subcellular distribution in the
kidney would establish whether polycystin-2 may function as an
intracellular calcium release channel in vivo. We fractionated whole
kidneys from C57BL/6J mice on iodixanol-based linear density gradients and compared the distribution of polycystin-2 with the distribution of known proteins from the plasma membrane, Golgi
apparatus and ER (Fig. 5). The distribution of polycystin-2 on the
gradient was indistinguishable from that of the ER membrane protein calnexin, and differed significantly from the distribution of
epidermal growth factor receptor (EGFR) and sodium proton
exchanger 3 (NHE3), a pair of proteins that are resident within
basolateral and apical plasma membranes, respectively (Fig. 5a).
The distribution of polycystin-2 also differed from that of the Golgi
58K protein (Fig. 5a). As the loading of the immunoblots was
quantitative (see Methods), the relative intensity of the anti-polycystin-2 immunoreactive band in each lane is representative of the
absolute steady state amount of polycystin-2 in each fraction
(Fig. 5a). The vast majority of polycystin-2 protein in the kidney
was excluded from the plasma membrane fractions altogether.
Small amounts of polycystin-2 and calnexin did overlay the region
of the gradient containing plasma membrane and Golgi proteins.
The identical relative distribution for the known ER-restricted protein calnexin suggests that this finding results from an inherent limitation of the continuous fractionation procedure.
To refine further the biochemical data for the subcellular localization in the ER compartment, we examined the sensitivity of
polycystin-2 in each fraction to Endo H. Typically, proteins that
remain sensitive to Endo H are localized to the ER and cis portions
of the Golgi apparatus. We have shown previously that wild-type
polycystin-2 is confined to the ER and is completely sensitive to
Endo H, whereas truncated forms of the protein that traffic to the
plasma membrane acquire Endo H resistance12. In keeping with
these observations, we now find that polycystin-2 from all fractions
of kidney tissue retained sensitivity to Endo H (Fig. 5b,c). The partial Endo H resistance of a minute proportion of total kidney polycystin-2 in three fractions overlying the plasma membrane region
(fractions 4–6, Fig. 5b) resulted from glycosidase inhibitory factors
that were concentrated in this region of the gradient. When these
fractions were spin dialysed over a Centricon YM-100 membrane
to remove the lower molecular weight molecules, polycystin-2 in
the fractions became completely sensitive to Endo H (Fig. 5c).
Thus, using whole tissue fractionation and sensitive biochemical
tests, we show that polycystin-2 is an abundant ER membrane protein in native kidney tissue. We suggest that the ER may be the primary site of its physiological activity.
Discussion
Although the channel function of polycystin-2 was anticipated by
its primary structure3, implied by the activity of its homologue31,
and suggested in single channel studies16,17, the function of this
human disease gene in kidney epithelia has remained elusive. In the
present study, we provide evidence that polycystin-2 expressed in
the ER of epithelial cells is a functional calcium-activated high conductance channel that is permeable to divalent cations. Increased
levels of intracellular calcium activate polycystin-2-mediated
release of calcium from intracellular stores. Furthermore, a truncated polycystin-2 mutant retains channel activity, but is no longer
calcium-activated or voltage-dependent, and cannot function to
release calcium from intracellular stores. By contrast, a pathogenic
missense mutation18 results in loss of polycystin-2 channel activity
altogether. The missense mutant retains the subcellular localization
and C-terminal-mediated protein interaction and regulatory
domains of the wild-type protein, thus providing evidence that the
loss of channel function alone is sufficient to cause polycystic kidney
NATURE CELL BIOLOGY ADVANCE ONLINE PUBLICATION http://cellbio.nature.com
© 2002 Macmillan Magazines Ltd
5
articles
disease. Polycystin-2 also shows voltage dependence in lipid bilayers. Although direct measurement of ER membrane potentials has
not been reported and the in vivo significance of this finding is
uncertain, it remains possible that localized membrane potentials
in the 0–10 mV range, sufficient to inactivate polycystin-2, may
occur through calcium release from the ER28. Overall, the current
findings suggest that in epithelial cells, polycystin-2 functions
through the regulated release of calcium from intracellular stores.
The loss of this intracellular calcium signalling mechanism results
in the defect associated with polycystic kidney disease.
In keeping with its function as an intracellular calcium release
channel, tissue fractionation studies show that polycystin-2 is
expressed abundantly in the ER of native kidney tissue. The absence
of any detectable polycystin-2 with resistance to Endo H digestion
supports the hypothesis that polycystin-2 does not traffic past
medial Golgi stacks, and therefore is not inserted into the plasma
membrane. Findings in a recent study of Chinese hamster ovary
(CHO) cells transfected with polycystin-1 and -2 have been interpreted as showing that polycystin-1 is required to co-assemble a
polycystin complex and bring polycystin-2 to the cell surface9.
However, direct evidence that polycystin-2 is actually inserted into
the plasma membrane is lacking, and, although the authors found
a novel surface cation channel activity when co-expressing polycystin-1 and -2, they were unable to demonstrate that polycystin-2
comprised the channel pore9. By contrast, the C. elegans orthologues of polycystin-1 and -2, location of vulva (LOV)-1 and PKD2 also function in a common pathway, but localization of PKD-2 is
not dependent on LOV-1 function10. Furthermore, in C. elegans,
functional localization of the polycystin-2 orthologue is limited to
intracellular membranes and the membrane of sensory cilia, and is
not detected at the plasma membrane of the cell body or axon10.
Interestingly, the only other instance of isoforms of an ER calcium
channel existing in a cell surface membrane has been the InsP3R in
the sensory cilia of olfactory neurons in crustaceans32,33. Although
we did not detect any evidence of polycystin-2 in the monocilia of
renal tubular cells, our study cannot exclude the presence of a
minute fraction of cellular polycystin-2 in these structures.
Calcium release activity mediated by polycystin-2 did not
require the co-expression of exogenous polycystin-1 in LLC-PK1
cells. This finding, in conjunction with the bilayer data, suggests
that polycystin-2 channel function does not require co-assembly
with polycystin-1, although a function for low levels of native polycystin-1 in these cells cannot be excluded. The absence of intracellular calcium release in cell expressing the PKD2-D511V mutant,
which retains the protein interactions of wild-type polycystin-2,
argues strongly that polycystin-2 itself forms the channel. It
remains possible, however, that channel activity requires association with other proteins. Polycystin-1 does associate with polycystin-2 (refs 8,9) and it is likely to be integral to the regulation of
the polycystin-2 activity in vivo10. The interaction between these
two proteins may occur either as polycystin-1 is transported
through the ER and early Golgi compartments, or it may occur
across closely apposed plasmalemmal and ER membranes near the
cell surface34. The paradigm for the association and functional
interdependence of channel proteins in the plasma and ER membranes, L-type calcium channels and the RyR, respectively, is the
basis of excitation-contraction coupling in muscle cells35. This conformational coupling model has been extended to nonexcitable cell
types, suggesting that the coupling of ER calcium release channels
to cell surface channel receptors may be a generalized principle in
nature36,37. Indeed, the polycystin-1 and polycystin-2 receptorchannel complex may fit this paradigm34. The polycystin-1dependent channel activity at the cell surface9 may result in local
increases in intracellular calcium that are sufficient to stimulate the
release of calcium from intracellular stores through closely associated polycystin-2 channels in the ER.
Localized increases in intracellular calcium mediate a variety of
subcellular processes. These processes include targeted fusion of
6
specialized cytosolic vesicles with discrete domains in the plasma
membrane or activation of kinase cascades that stimulate cell-type
specific gene expression38. Either or both of these general functions
could subsume the function of the polycystin pathway. Known cellular functions of polycystin orthologues include the acrosome
reaction, a calcium-regulated membrane fusion event in sea urchin
sperm39,40, mechanosensory neuron dependent male mating behaviours in C. elegans10,41, and recently, a defect in the late stages of
endocytosis in mucolipidosis type IV42,43. Based on our current
results, further studies into the function of intracellular signals and
their regulation in various cell models will, in aggregate, yield crucial insights into the pathogenesis of polycystic kidney disease.
Methods
Cell lines, cell culture and vesicle preparation
LLC-PK1 cell lines expressing wild-type PKD2 (GenBank Accession number U50928), the PKD2L703X mutant and vector controls have been previously described12. The PKD2-D511V mutant cDNA
construct was made by introducing GTT (V) to replace GAT (D) at codon 511 by site-directed mutagenesis in pCDNA3.1. Cells stably overexpressing PKD2-D511V were obtained by selection with
Zeocin (Invitrogen, Carlsbad, CA) according to manufacturer’s instruction. Expression of the mutant
protein was evaluated by immunoblotting and immunofluorescence cell staining12. For the tetracycline-inducible clone, LLC-PK1 cells were transfected with pUHD15-1Neo (a gift from H. Bujard,
Heidelberg, Germany) encoding a tetracycline-dependent transactivator. Clone LtTA-2.22 showed
~65-fold induction after transient transfection and was chosen to establish stable clones. PKD2-HA
(ref. 12) in the expression plasmid pUHD10-3 (gift from H. Bujard) was transfected into LtTA-2.22
along with the selection plasmid pBabe-Puro. Overexpression of polycystin-2 in LtTA-2.22 cells was
inhibited by the addition of doxycyline (10 ng ml−1) to the culture medium. Membrane vesicles
enriched for ER from LLC-PK1 cells were isolated by differential centrifugation in the presence of protease inhibitors 44. The expression of polycystin-2 in vesicles was monitored by immunoblotting with
YCC2 polyclonal antisera12.
Single channel recordings
Vesicles enriched in ER membranes were fused to lipid bilayers containing phosphatidylethanolamine
and phosphatidylserine (3:1 w/w; Avanti Polar Lipids, Alabaster, AL) dissolved in decane (40 mg lipid
ml−1)23. A potassium chloride gradient, with 600 mM potassium chloride on the side of vesicle incorporation (cis side) and 0 mM potassium chloride on the opposite side (trans side), facilitated fusion.
Experiments were performed with 250 mM HEPES-Tris at pH 7.35 on the cis side and 250 mM HEPES
with either 55 mM Ba(OH)2, 55 mM Ca(OH)2, or 55 mM Mg(OH)2, at pH 7.35, on the trans side of
the bilayer. In some experiments, we determined the orientation of ER membrane proteins in the lipid
bilayer through activation of the RyR. 2-APB (100 µM) was applied to the cytosolic side of the protein
to test its effect on the polycystin-2 channel. Experiments were recorded under voltage-clamp conditions. The trans-side was ground (reference electrode), the cis-side was the input electrode. Using cellular conventions, calcium flux from the ER lumen into the cytoplasm was defined as an inward current,
where inward currents are shown as downward deflections. Data were filtered at 1 kHz and digitized at
3 kHz, directly transferred to a computer and analysed with pClamp version 6.0.3 software (Axon
Instruments Burlingame, CA). The concentration of free calcium on the cytosolic side of the protein
(cis-side of the lipid bilayer) was adjusted as described45. In each experiment, the voltage-dependence
of the polycystin-2 current was assessed and most channels were recorded under suboptimal activation
conditions, by applying a holding potential of −3 to −5 mV.
Optical recordings of intracellular calcium concentrations
Wild type and stably transfected LLC-PK1 cells were grown on glass coverslips to subconfluent density.
During experiments, cells were kept at room temperature in a perfusion chamber on the microscope
stage. Cells were superfused continuously with extracellular solution (ECS) at 1 ml min−1. ECS contained 137 mM sodium chloride, 5 mM potassium chloride, 2 mM calcium chloride, 1 mM Na2HPO4,
1 mM magnesium sulphate, 10 mM HEPES and 22 mM glucose at pH 7.4. Cells were incubated in
ECS containing 4 µM cell-permeant fluo-3 (Molecular Probes, Eugene, OR) with 0.04% DMSO and
0.02% pluronic acid for 15–30 minutes and washed in ECS before optical recording. Drugs applied to
the bath solution were: thapsigargin (1–10 µM; Calbiochem-Novabiochem, San Diego, CA), vasopressin (0.07–70 µM; Calbiochem-Novabiochem), 2-APB (10–200 µM; Sigma, St Louis, MO). Fluo-3
fluorescence in loaded cells was measured with a BioRad MRC-1024 system (Biorad, Hercules, CA).
Cells were selected randomly before each experiment and the borders of the entire cell were outlined
manually to eliminate any selective contribution from nonuniform distribution of the dye within cells.
Dye-loaded cells were excited at 488 nm and increases in intracellular calcium were measured at an
emission wavelength of 522 nm. Images were acquired every 500 ms. Changes in fluorescence intensity
were expressed as a ratio, F/F0, of fluorescence intensity during drug application (F) and mean baseline
fluorescence intensity (F0). Transient durations were measured from the start of the calcium transients
until fluorescence changes returned to the average prestimulus background. Nonstimulus-related,
spontaneously occurring changes in fluorescence intensity, as well as changes after the application of
control substances, were in the range of 1–5% of F/F0.
Tissue fractionation, immunoblotting and glycosylation analysis
Tissue homogenates were prepared from eight-week-old C57BL/6J mouse kidneys. Kidneys were
homogenized at 4 °C in buffer containing 10 mM Tris at pH 7.5, 120 mM sodium chloride, 20 mM
potassium chloride, 1 mM EGTA, 1 mM EDTA and protease inhibitor cocktail (2 mm PMSF, 10 µg
ml−1 leupeptin, 10 µg ml−1 pepstatin A and 1 µg ml−1 aprotonin). A postnuclear supernatant (PNS) was
obtained by centrifugation at 1,000g for 10 min and a microsome-rich supernatant (S20) was obtained
by centrifugation of the PNS at 20,000g for 20 min. 2 ml of the S20 supernatant was layered onto
NATURE CELL BIOLOGY ADVANCE ONLINE PUBLICATION http://cellbio.nature.com
© 2002 Macmillan Magazines Ltd
articles
preformed, precooled 8–34% linear Optiprep gradients (Life Technologies, Grand Island, NY) and
centrifuged for 15 h at 100,000g. 18 equal-volume fractions (2 ml each) were collected from the top to
the bottom of the gradient. Total protein from each fraction was precipitated with trichloracetic acid
and pellets were solubilized in equal volumes of Laemmli sample buffer. Equal sample volumes were
separated by 7.5% SDS–polyacrylamide gel electrophoresis (PAGE) gels and transferred to PVDF
membranes. Primary antibodies for immunoblotting were : anti-polycystin-2 (1:8,000; YCC2 (ref. 12)),
anti-calnexin (1:2,000; Stressgen Biotechnologies Corporation, Victoria, BC); anti-EGFR (1:400; Signal
Transductional Laboratories, Lexington, KY (Catalogue number E12020)), anti-NHE3 (1:500;
Chemicon International, Temecula, CA (Catalogue number mAb3134)); anti-Golgi 58K protein
(1:5,000; Sigma (Catalogue number G2404)). For glycosylation analysis, 500-µl aliquots of individual
fractions were digested overnight at 37 °C with 2500U of Endo H, as recommended by manufacturer’s
instructions (New England Biolabs, Beverly, MA). For controls, equal aliquots of each fraction were
treated identically, except that no Endo H was added to the incubation. To remove potential Endo H
inhibitory factors from selected fractions, the fractions were centrifuged through Centricon YM-100
columns (Millipore, Bedford, MA) at 1,000g for 30 min at 4 °C before Endo H digestion.
RECEIVED 11 OCTOBER 2000; REVISED 13 NOVEMBER 2001; ACCEPTED 3 DECEMBER 2001;
PUBLISHED 18 FEBRUARY 2002.
1. Gabow, P. A. & Grantham, J. J. Diseases of the kidney. Schrier,R. W. & Gottschalk,C. W. (eds),
521–560 (Little, Brown; Boston, 1997).
2. The European Polycystic Kidney Disease Consortium. The polycystic kidney disease 1 gene encodes
a 14 kb transcript and lies within a duplicated region on chromosme 16. Cell 77, 881–894 (1994).
3. Mochizuki, T. et al. PKD2, a gene for polycystic kidney disease that encodes an integral membrane
protein. Science 272, 1339–1342 (1996).
4. Qian, F., Watnick, T. J., Onuchic, L. F. & Germino, G. G. The molecular basis of focal cyst formation
in human autosomal dominant polycystic kidney disease type I. Cell 87, 979–987 (1996).
5. Wu, G. et al. Somatic inactivation of Pkd2 results in polycystic kidney disease. Cell 93, 177–188
(1998).
6. Lu, W. et al. Perinatal lethality with kidney and pancreas defects in mice with a targeted Pkd1
mutation. Nature Genet. 17, 179–181 (1997).
7. Wu, G. et al. Cardiac defects and renal failure in mice with targeted mutations in Pkd2. Nature
Genet. 24, 75–78 (2000).
8. Qian, F. et al. PKD1 interacts with PKD2 through a probable coiled-coil domain. Nature Genet. 16,
179–183 (1997).
9. Hanaoka, K. et al. Co-assembly of polycystin-1 and -2 produces unique cation-permeable currents.
Nature 408, 990–994 (2000).
10. Barr, M. M. et al. The Caenorhabditis elegans autosomal dominant polycystic kidney disease gene
homologs lov-1 and pkd-2 act in the same pathway. Curr. Biol. 11, 1341–1346 (2001).
11. Littleton, J. T. & Ganetzky, B. Ion channels and synaptic organization: analysis of the Drosophila
genome. Neuron 26, 35–43 (2000).
12. Cai, Y. et al. Identification and characterization of polycystin-2, the PKD2 gene product. J. Biol.
Chem. 274, 28557–28565 (1999).
13. Markowitz, G. S. et al. Polycystin-2 expression is developmentally regulated. Am. J. Physiol. 277,
F17–F25 (1999).
14. Ong, A. C. et al. Coordinate expression of the autosomal dominant polycystic kidney disease proteins, polycystin-2 and polycystin-1, in normal and cystic tissue. Am. J. Pathol. 154, 1721–1729
(1999).
15. Obermuller, N. et al. The rat Pkd2 protein assumes distinct subcellular distributions in different
organs. Am. J. Physiol. 277, F914–F925 (1999).
16. Gonzalez-Perrett, S. et al. From the Cover: Polycystin-2, the protein mutated in autosomal dominant polycystic kidney disease (ADPKD), is a Ca2+-permeable nonselective cation channel. Proc.
Natl Acad. Sci. USA 98, 1182–1187 (2001).
17. Vassilev, P. M. et al. Polycystin-2 is a novel cation channel implicated in defective intracellular
ca(2+) homeostasis in polycystic kidney disease. Biochem. Biophys. Res. Commun. 282, 341–350
(2001).
18. Reynolds, D. M. et al. Aberrant splicing in the PKD2 gene as a cause of polycysitic kidney disease. J.
Am. Soc. Nephrol. 10, 2342–2351 (1999).
19. Lehtonen, S. et al. In vivo interaction of the adapter protein CD2-associated protein with the type 2
polycystic kidney disease protein, polycystin-2. J. Biol. Chem. 275, 32888–32893 (2000).
20. Torres, V. E. et al. Vascular expression of polycystin-2. J. Am. Soc. Nephrol. 12, 1–9 (2001).
21. Tinker, A. & Williams, A. J. Divalent cation conduction in the ryanodine receptor channel of sheep
cardiac muscle sarcoplasmic reticulum. J. Gen. Physiol 100, 479–493 (1992).
22. Bezprozvanny, I. & Ehrlich, B. E. Inositol (1,4,5)-trisphosphate (InsP3)-gated Ca channels from
cerebellum: conduction properties for divalent cations and regulation by intraluminal calcium. J.
Gen. Physiol. 104, 821–856 (1994).
23. Ehrlich, B. E. & Watras, J. Inositol 1,4,5-trisphosphate activates a channel from smooth muscle sarcoplasmic reticulum. Nature 336, 583–586 (1988).
24. Harteneck, C., Plant, T. D. & Schultz, G. From worm to man: three subfamilies of TRP channels.
Trends Neurosci. 23, 159–166 (2000).
25. Tsien, R. W., Lipscombe, D., Madison, D. V., Bley, K. R. & Fox, A. P. Multiple types of neuronal calcium channels and their selective modulation. Trends Neurosci. 11, 431–438 (1988).
26. Bezprozvanny, I. & Ehrlich, B. E. The inositol 1,4,5-trisphosphate (InsP3) receptor. J. Membr. Biol.
145, 205–216 (1995).
27. Zuhlke, R. D. & Reuter, H. Ca2+-sensitive inactivation of L-type Ca2+ channels depends on multiple cytoplasmic amino acid sequences of the alpha1C subunit. Proc. Natl Acad. Sci. USA 95,
3287–3294 (1998).
28. Tinker, A., Lindsay, A. R. & Williams, A. J. Cation conduction in the calcium release channel of the
cardiac sarcoplasmic reticulum under physiological and pathophysiological conditions. Cardiovasc.
Res. 27, 1820–1825 (1993).
29. Watras, J., Bezprozvanny, I. & Ehrlich, B. E. Inositol 1,4,5-trisphosphate-gated channels in cerebellum: presence of multiple conductance states. J. Neurosci. 11, 3239–3245 (1991).
30. Dibas, A. I., Rezazadeh, S. M., Vassan, R., Mia, A. J. & Yorio, T. Mechanism of vasopressin-induced
increase in intracellular Ca2+ in LLC-PK1 porcine kidney cells. Am. J. Physiol. 272, C810–C817
(1997).
31. Chen, X. Z. et al. Polycystin-L is a calcium-regulated cation channel permeable to calcium ions.
Nature 401, 383–386 (1999).
32. Munger, S. D. et al. Characterization of a phosphoinositide-mediated odor transduction pathway
reveals plasma membrane localization of an inositol 1,4,5-trisphosphate receptor in lobster olfactory receptor neurons. J. Biol. Chem. 275, 20450–20457 (2000).
33. Cadiou, H. et al. Basic properties of an inositol 1,4,5-trisphosphate-gated channel in carp olfactory
cilia. Eur. J. Neurosci. 12, 2805–2811 (2000).
34. Somlo, S. & Ehrlich, B. Human disease: calcium signalling in polycystic kidney disease. Curr. Biol.
11, R356–R360 (2001).
35. Beam, K. G. & Franzini-Armstrong, C. Functional and structural approaches to the study of excitation-contraction coupling. Methods Cell Biol. 52, 283–306 (1997).
36. Kiselyov, K. et al. Functional interaction between InsP3 receptors and store-operated Htrp3 channels. Nature 396, 478–482 (1998).
37. Kiselyov, K., Mignery, G. A., Zhu, M. X. & Muallem, S. The N-terminal domain of the IP3 receptor
gates store-operated hTrp3 channels. Mol. Cell 4, 423–429 (1999).
38. Berridge, M. J., Bootman, M. D. & Lipp, P. Calcium—a life and death signal. Nature 395, 645–648
(1998).
39. Moy, G. W. et al. The sea urchin sperm receptor for egg jelly is a modular protein with extensive
homology to the human polycystic kidney disease protein, PKD1. J. Cell Biol. 133, 809–817 (1996).
40. Neill, A. T., Moy, G. W. & Vacquier, V. D. Characterization of sea urchin polycystin-2. Mol. Biol. Cell
11, 2105 (2000).
41. Barr, M. M. & Sternberg, P. W. A polycystic kidney disease gene homolog required for male mating
behavior in Caenorhabiditis elegans. Nature 401, 386–389 (1999).
42. Bassi, M. T. et al. Cloning of the gene encoding a novel integral membrane protein, mucolipidinand identification of the two major founder mutations causing mucolipidosis type IV. Am. J. Hum.
Genet. 67, 1110–1120 (2000).
43. Sun, M. et al. Mucolipidosis type IV is caused by mutations in a gene encoding a novel transient
receptor potential channel. Hum. Mol. Genet. 9, 2471–2478 (2000).
44. Kim, D. H., Ohnishi, S. T. & Ikemoto, N. Kinetic studies of calcium release from sarcoplasmic reticulum in vitro. J. Biol. Chem. 258, 9662–9668 (1983).
45. Fabiato, A. Computer programs for calculating total from specified free or free from specified total
ionic concentrations in aqueous solutions containing multiple metals and ligands. Methods
Enzymol. 157, 378–417 (1988).
ACKNOWLEDGEMENTS
We thank M. Caplan, W. Echevarria and M. Nathanson for helpful discussions. We also thank P.
Aronson and E.Thrower for critical reading of the manuscript; Z. Huang for the LtTA-2.22 clone, A.
Cedzich for the stably transfected LtTA-2.22 cell lines and B. DeGray for technical assistance. This
work was supported by grants from the National Institutes of Health to S.S. (DK57328) and B.E.E.
(DK57328, GM51480) and from the American Heart Association to L.G. (9730148N) and Y.C.
(0130207N). P.K. and Y.M. were supported by National Kidney Foundation Research Fellowship
Awards. P.K. was supported by a BASF Scholarship. The authors are investigators in the Yale Center for
the Study of Polycystic Kidney Disease (P50 DK57328).
Correspondence and requests for material should be addressed to S.S
NATURE CELL BIOLOGY ADVANCE ONLINE PUBLICATION http://cellbio.nature.com
© 2002 Macmillan Magazines Ltd
7